The treatment of advanced colorectal cancer offers an ideal environment for the development and evaluation of pharmacogenetics. The availability of multiple therapeutic options with clear antitumor activity requires clinical decisions to be made for individual patients and these decisions are best made with quantitative information. However, there are currently insufficient tools to guide the selection of therapy for advanced colorectal cancer. The recent FDA approval of irinotecan (CPT-11) and oxaliplatin (OXAL) for advanced colorectal cancer, in addition to the historical standard 5-fluorouracil (5FU), has led to the development of several active combination chemotherapy regimens for this disease. Indeed, the recently completed Gastrointestinal Intergroup study N9741 compared 5FU/CPT-11, 5FU/OXAL, and CPT-11/OXAL, with observed objective response rates of 28-38%. The previous reports of genetic variants associated with toxicity and/or activity from these three agents provides a promising approach for the development of a pharmacogenetic strategy to select therapy for advanced colorectal cancer. Therefore, this project will address the following Specific Aims: 1. Determine the predictive impact of genetic variants in candidate genes on severe toxicity from 5FU, CPT-11, or OXAL therapy in 575 patients from the N9741 study. 2. Define the association of genetic variants in candidate genes with response to therapy, time to progression, overall survival, and quality of life after 5FU, CPT-11, or OXAL regimens for advanced colorectal cancer. 3. Develop methods for incorporating Genetic Polymorphism Profiles into Decision Making. These aims provide a comprehensive framework for our strategy to provide the first prospective information on the integration of pharmacogenetic information into the selection of therapy for advanced colorectal cancer.